This study continues longitudinal research designed to strengthen the empirical basis for assessment, taxonomy, epidemiology, and prediction of psychopathology in children, adolescents, and young adults. It will reassess at ages 10 to 22 a unique national sample that was previously assessed via home interviews and standardized ratings in 1986 at ages 4 to 16 and 1989 at ages 7 to 19. The existing assessment data include multisource (parent, teacher, self) reports of behavioral/emotional problems and competencies, family composition, family mental disorders, socioeconomic status, child care, arrangements, welfare support, and stressful events occurring between 1986 and 1989. The 1989 assessment also included the following signs of disturbance: (a) problem scores in the clinical range; (b) referral for mental health services; (c) referral for special education; (d) school behavior problems, suspension, expulsion; (e) trouble with police, arrest, adjudication; (f) suicide attempts; (g) parent's judgment that the child needed professional help; and a composite index of items (a)-(g). The 1989 follow-up obtained data for 90.7% (N = 2,479) of the subjects originally assessed in 1986. By reassessing the 2,479 subjects assessed in 1986 and 1989, the proposed 6-year follow-up is designed to: (1) test the hypothesis that practical, standardized measures of parent-, teacher-, and self-reported problems and competencies will predict the signs of disturbance listed above, plus substance abuse; (2) test the ability of family variables and intervening stresses to predict the signs of disturbance and problem scores; (3) compare and model predictions of the signs of disturbance from optimal combinations of family variables, subject variables, and intervening events; (4) test the differential incidence and cumulative prevalence hypotheses for explaining higher rates of psychopathology among lower than upper socioeconomic groups; (5) advance the taxonomic basis for diagnosis of child/adolescent disorders by testing the stability, predictive power, discriminative power, and correlates of 8 empirically based syndrome constructs over 3- and 6-year periods; (6) compare the ability of the 8 syndrome constructs to predict outcomes; (7) track the development of psychopathology across the crucial transitional period from adolescence to adulthood.